Getty Images/iStockphoto

Sisu adds new decision intelligence capabilities

The startup aims to augment human decision-making with automation, and its new tools find and explain changes in data, analyze their effect and prescribe what to do next.

Startup analytics vendor Sisu, guided by the concept of decision intelligence, added new capabilities designed to augment the decision-making process with automation.

Most recently, on Aug. 22 at the Gartner Data and Analytics Summit in Orlando, Fla., the San Francisco-based vendor unveiled four new capabilities, including ones that let organizations detect trends and anomalies, monitor metrics at enterprise scale and predict the impact of changes in metrics.

In addition, the vendor -- founded in 2018 -- revealed new integrations with data management platforms, including Databricks, and analytics platforms such as Tableau and Looker.

All are now generally available.

A complement to BI

Decision intelligence is not BI, nor is it designed to replace BI, according to Sisu founder and CEO Peter Bailis. Instead, it's meant to complement BI.

BI is the process of analyzing data to make decisions. But the amount of data organizations collect now has exploded in volume and is increasing in complexity.

It's become too much for traditional BI, which relies on humans looking at their organization's data and deriving the insights that inform decisions. It's simply not possible for even a team of data experts to monitor every data set, notice every change and react in real time.

So decision intelligence -- the use of augmented intelligence and machine learning to inform automation -- is needed, according to Bailis.

"There's this realization that BI is really good at manually driven questions and answers," he said. "But to really get to the next level where you have data and need help getting answers from the data automatically, you need more than a BI tool."

The cloud has only complicated matters, but not in a bad way, Bailis continued.

The cloud lets organizations connect to more data sources and ingest data more quickly. That, in turn, leads to better informed decisions. But it also increases the workload on data teams and makes finding the right data for a given decision more complicated.

"Decision intelligence is really empowering that decision-maker who may have a team of analysts that is overloaded and wants answers 24 hours a day, seven days a week," Bailis said. "There's no human that can keep up with that. Decision-making is still a human process, but the algorithms take all the different data dimensions and tell the human which one or two they should look at."

Ultimately, decision intelligence is a concept that should result in increased productivity, according to Wayne Eckerson, founder and principal consultant at Eckerson Group.

By saving data analysts the copious amount of time it takes to identify the right data source -- or sources -- for a given project or decision, automation helps those analysts get to the decision-making process much faster and lets the organization act quicker.

"[Decision intelligence capabilities] won't replace BI and data science tools, but they can make analysts more productive by doing the heavy lifting of sifting through large volumes of data and identifying the key issues that the analysts need to focus on," Eckerson said. "It makes analysts more productive by pointing them in the right direction, shortening the time to insight and action."

New capabilities

Decision intelligence is really empowering that decision-maker who may have a team of analysts that is overloaded and wants answers 24 hours a day, seven days a week. There's no human that can keep up with that.
Peter BailisFounder and CEO, Sisu

Toward the end of enabling decision intelligence and its goal of making analysts more efficient and shortening the time needed to derive insights and act, Sisu launched its most recent new capabilities and integrations.

From its inception, the vendor aimed to automatically monitor an organization's data for changes and anomalies, discover why that organization's metrics changed and anomalies arose, and provide an explanation. As time passed, Sisu realized it also needed to tell organizations what to do about those changes and anomalies.

Taken together, the four new capabilities enhance Sisu's ability to discover why metrics change and anomalies arise, inform organizations what to do about changes and anomalies, and predict the future effect of those changes and anomalies.

The capabilities include the following:

  • the automatic monitoring of metrics and detection of trends and anomalies in data at enterprise scale with automatic alerts sent to data stakeholders via email, Slack and other tools;
  • the enhanced use of AI and machine learning to reduce bias and risk in diagnosing the reasons behind changes in metrics and existence of anomalies;
  • the ability to predict the potential effects of changes in metrics with smart charts and ranked results so organizations can better plan for the future; and
  • direct integrations with data management and analytics platforms -- including Databricks, DBT Labs, Tableau and Looker -- so organizations can take action based on the insights derived from Sisu.

"Decision-making isn't just about the why," Bailis said. "It's a continuous process that includes knowing what is happening, why it's happening and what to do about it. Now, we have generally available capabilities that do all three, including our first foray into being able to do predictions. These updates complete the loop."

Even so, some of Sisu's new capabilities are also provided on other analytics vendors' platforms , according to Eckerson. In particular, anomaly detection and integrations that enable in-database work with data are relatively common.

Other new capabilities, however, are advanced.

"Some of [the new capabilities] are catch-up," Eckerson said. "Some is good new functionality. I like the prediction and action features a lot since that is where these tools need to head."

Going forward, Eckerson said he'd like to see Sisu add even more tools that let organizations act based on insights provided by the vendor's decision intelligence capabilities. Currently, the ability to act requires a human decision based on the push notifications they receive.

"The real fun is when the actions can automatically update things, which I'm sure is coming soon," Eckerson said.

A sample Sisu screenshot
A sample Sisu screenshot displays trends related to an organization's churn rate.

Future plans

As Sisu currently exists, it's not designed to be a standalone analytics platform, according to Bailis.

Its decision intelligence capabilities complement traditional BI platforms such as Tableau, Qlik and Microsoft Power BI rather than replace them. Sisu lacks some of the basic functionality of traditional BI platforms, including building dashboards and developing data visualizations like pie charts and heat maps.

However, that's where the vendor is headed and what it aims to be as it grows. Its total funding to date is about $130 million, and it still has fewer than 100 employees, according to Crunchbase.

"We have ambitions over the next several years of becoming a primary tool," Bailis said. "Today, what we do is step in and complement in ways where an organization can't hire and scale up to have enough analysts."

Beyond adding some of the basic functionality it will take to eventually become an organization's primary analytics platform, Sisu plans to add algorithms that increase automation and improve its decision intelligence capabilities.

Sisu plans to add features that will let an organization act on an insight with one click and add more capabilities for scenario planning that help organizations understand which potential action is best.

"We want to be more powerful and more accurate," Bailis said.

Dig Deeper on Business intelligence technology

Data Management
Content Management